Are Point-of-Care devices suitable for the management of Dengue Virus infections? A comparison with mainframe analysers using External Quality Assurance (EQA) data

Abstract

Introduction: Dengue virus (DV) is an arthropod-borne disease that has a significant impact on human health and the economy. Globally, around 2.5 billion individuals have been affected by the virus. Early and rapid dengue identification is crucial for monitoring transmission dynamics and enabling swift responses to outbreaks. Mainframe immunoassay analysers are commonly used to detect Dengue Virus (DV) serology; however, point-of-care devices enable front-line healthcare workers to optimise limited resources and reduce the economic burden of this widespread disease. The RCPAQAP has a Dengue Virus Serology EQA program covering DV IgG, IgM and NS1 Ag markers suitable for both mainframe and Point of Care (PoC) assays.We sought to evaluate the diagnostic accuracy of PoC devices compared to mainframe ELISA/FIA assays based on the returned EQA results. Method: Clinical serum/plasma samples from confirmed positive/negative DV patients were sourced from Australian and overseas laboratories. The acute NS1 Antigen (NS1 Ag) positive samples were produced by spiking NS1 Ag (type 4) inactivated cell culture fluid into negative serum samples. A total of 20 samples (4 per year) representing various disease phases (based on pre-testing data), 5 acute, 5 convalescent, 2 recovered, 1 indeterminant and 7 known negatives, were dispatched to an average of 110 participating laboratories over a five-year study period (2019-2023). Quantitative and qualitative (positive/negative/equivocal) results from mainframe analysers were compared to the qualitative (positive/negative) PoC results using RCPAQAP software. Results: The acute and convalescent samples achieved >80% positive consensus for DV IgG and IgM across all mainframe analysers and PoC devices, except for 1 convalescent sample which had a 78% positivity for PoC. The recovered specimens returned ≥90% positivity for IgG for all mainframe methods, and 53% for the PoC methods. The indeterminate sample was negative for IgG across all methods, the IgM for this sample was 56% positive for IgM on mainframe, but >90% negative on PoC. NS1 returned >80% positive for acute samples on mainframe and between 53 and 75% for PoC. The NS1 was >80% negative for recovered, convalescent and indeterminate samples for all methods. There was >95% consensus for all methods for the negative samples. Discussion: Based on this study, all DV IgG methods provided suitable results to identify and manage DV infections during the acute and convalescent phases. PoC methods were less sensitive for detection of IgM and NS1 Ag for the acute/convalescent samples and for IgG in the recovered. The >95% agreement for the true negatives confirms an acceptable specificity for all methods. Conclusion: While mainframe analysers and PoC demonstrated acceptable performance in this EQA review, awareness of the variation in sensitivity and specificity between methods for patients who present in different stages of their infection is important when interpreting results.